The deal is that score calculated per shard using different statistical metrics. From another side we have a relevance – at ES that is a number (it is called score) that reflects how good according document matches our search. Index data is spread over different cluster nodes using special building blocks that are called shards. Here I will try to explain what it means as short as possible. If you are interested in details now – please apply to the 5th udemy course section at shards lecture (the link to the course is at the bottom of current article). I don’t want to put at your head to much information at once.
For testing purposes we have to create an index with just one primary shard. Ok, seems all is good but there is one essential note.
